#NoEstimates is to a large extent a reaction to those businesses where developers are asked to make instant, off-the-cuff estimates of how much time/effort/money is required to do something, these estimates are used to make concrete plans, and developers are then chastised when their estimates turn out to be false. Making proper estimates with realistic upper and lower limits and probabilities takes time, and therefore money. If you are in an environment where people are afforded the time to make these estimates, and it is worth it in practice, then this is good. Not everyone works in such an environment. When people say "No Estimates", they mean "no off-the-cuff estimates that we'll be held to", and "no time consuming careful estimates". Instead, developers from the "no estimates" community try alternative ways of doing things. They often include implicit estimates ("small enough to work on" vs "too big, needs splitting", for example), but not hard-and-fast ones like the above. If a really coarse grain estimate of maximum cost is enough to make a decision then we don't need the effort of "proper" estimates. e.g. if the upside is $1 million, and a coarse grain estimate says less than 6 months, less than $500000, it's worth it. We don't need to spend the time (probably weeks) doing a "proper" estimate only to find it was "3-4 months". The time spent doing the detailed estimate is just waste.