Long time player & reader here, just joined. This started as scratch notes to help organize my thoughts, but I felt it may be of use to others so I cleaned it up a bit to post here. My hope is that this starts a constructive discussion and perhaps even helps you decide whether to harvest/remove or not. Terminology Harvest Resource: A builder can use a charge to harvest a resource to provide an instant significant but one-time only boost to the city that owns the tile, removing the resource in the process. Only bonus resources can be harvested. [Footnote: The term harvest can be confusing because it is often used to refer to the act of improving a resource (like with farm or plantation or similar), rather than removing the resource for a one-time gain.] Remove Feature: A builder can use a charge to remove a terrain feature to provide an instant significant but one-time only boost to the city that owns the tile, removing the feature in the process. Only three terrain features can be removed. The type of boost the city receives when a builder harvests a resource depends on the resource: Bananas, Cattle, Fish, Rice, Sheep, Wheat yield food. Deer, Stone yield cogs. Copper, Crabs yield gold. Similarly, the type of boost the city receives when a builder removes a feature depends on the feature: Marsh yields food. Woods yields cogs. Rainforest yields food and cogs. Everything stated below applies to these two builder actions on these resources and tile features only. Instant boost vs long-term benefit dilemma The question at hand is: should you opt for an instant significant one-time boost (henceforth called “instant boost”) at the cost of losing the benefits of the resource/feature & possible improvements (henceforth called “long-term benefit”), or should you not? Let's get a few things out of the way off the bat. Sometimes there are specific reasons why you want to choose one or the other: You need the tile for a district/wonder. You need the tile for a different improvement (for example to create three adjacent farms). You want to keep a resource / feature because a wonder you intend to build requires it (Stonehenge, Temple of Artemis, Mohabodhi Temple, Chichen Itza, Mont St. Michel). You want to keep a resource / feature because a district or wonder benefits from it (harbor, industrial zone, Temple of Artemis, Chichen Itza, Great Zimbabwe, Ruhr Valley). You want to get a eureka or inspiration. Timing constraints (for example, you need to harvest/remove right now to rush that wonder). Tile appeal and adjacency bonus concerns. Other modifiers such as your pantheon or Civ/leader bonuses. Looking at this far-from-exhaustive list, it becomes clear that the instant boost vs long-term benefit decision is a very complex one (and this complexity is a big reason why we all love this game!). Even without any specific reasons, choosing one over the other is quite difficult: the former can help you to get your empire off the ground fast but at the cost of leaving your lands barren; the latter can make your empire lag behind with a lot of catching up to do. Given this complexity, quoting exact figures is a futile attempt. Statements like “you remove a feature for x cogs at the cost of y for the builder charge, thus leaving you with a benefit of z=x-y...” attempt to quantify the trade-offs in a very simplistic manner. They fail to take into account how much of an impact this action has on your empire (for example, perhaps it helped you churn out two quick warriors to tip the scale and capture that neighboring city, which otherwise you would not have been able to, losing the benefits of that city, and snowballing from there). Rather, a better approach is to list recommendations or guidelines, with brief explanations. Some of these make a stronger case than others, so they are ordered in decreasing order of conviction. You can divide these into tiers, if you prefer. This list assumes that you do not have any specific reasons as outlined above, because there will always be situations that warrant exceptions. Guidelines Harvest/remove before building a district/wonder -- The resource and/or feature is removed automatically and the potential instant boost is lost otherwise. Establish Magnus in the city when choosing to harvest/remove -- This doubles the boost. Pursue ways to acquire builder charges cheaply -- Harvest/remove comes at the cost of a builder charge, making them very valuable (Pyramids, Liang, China LUA, Serfdom / Public Works); also applies to free builders (e.g., Aztec Eagle Warrior). Take advantage of production overflows -- When a builder performs the harvest/remove, the resulting instant boost reflects the benefits of active policies (for example, you may have Maritime Industries active, complete your galley instantly, and apply the remaining / overflow production to build a wonder/district next while benefiting from the additional 100% boost for the galley). Remove marsh and rainforest features -- They provide just +1 food without any options to improve upon (and also -1 appeal). Remove hill features and resources -- Hills can be improved with mines for +1 cogs (+1 with Apprenticeship and another +1 with Industrialization, plus other potential benefits) thus allowing you to reap the instant boost and offset the long-term benefit loss of the resource or feature with the long-term benefits of the mine improvement. Harvest/remove even if you already improved it, if warranted -- For example, you may have improved a hill stone tile for a eureka; once this is achieved, you can remove the improvement (no builder charge) and then harvest/remove. Yes, this incurs and additional builder charge because of the original improve and the eventual harvest/remove. Choose the Pantheon “Goddess of the Harvest” -- It provides faith equal to the boost gained. Choose the Dedication “Monumentality” -- The synergy with “Goddess of the Harvest” is just plain beautiful [In one game, I had Pyramids, Liang, Serfdom, and churned out over 40 builders on consecutive turns each with seven charges (I play at Marathon speed), and still had over 20,000 faith left when done (I even churned out six settlers in the process)! Needless to mention that I choose “Exodus of the Evangelists” the following era.]. Place your cities closer together (like 4 or 5 tiles apart rather than 6 or 7) -- You may be left with swathes of barren land otherwise (unless you have access to UIs, see next point). Pursue options for Civ/CS specific UIs (Ziggurat, Kurgan, …) -- Harvest/remove provides an instant boost but leaves barren land behind, and these offer options to put the barren tiles to use (much like mines do for hills). UIs are in general not too spectacular (in my opinion), but still better than nothing. Treat flat land woods and flat land bonus resources case by case -- These can be improved with lumber mill, farm, mine, quarry, plantation, camp, pasture, or fishing boat, each providing various different yields (some combination of cogs/food/gold). Since they are on flat land, there is no obvious improvement you can place on the barren tile (unless you have access to UIs, see #9). [Personally, I harvest most flat land bonus resources, except possibly some food yielding ones if the city needs food. I tend to keep some river flat land woods and improve with lumber mills (they provide up to +3 cogs, and I retain the option of removing the forest later on - flexibility is important). Consider changing your game start settings -- When starting a game, consider choosing Resources - Abundant (more resources), World Age - New (more hills and mountains), and Rainfall - Wet (more woods, rainforest, and marsh). This makes for a much more interesting gameplay where builders become much more valuable. Performing all harvest/remove actions on the very last turn is the optimal situation -- The instant boost you gain increases as your Civ progresses, so this maxes the instant boost; and the long-term benefit loss is zero, so this would be the optimal time. This is listed here just for completeness, of course, and just to get the message across that delaying harvest/remove improves the instant boost to long-term benefit loss ratio. And this is, of course, just from a simplistic point-of-view.