Predictions of forest succession have a tremendous influence on forest management decisions, strategies, and policies. Although generalized forest succession models are intensively studied, there is a distinct lack of long-term studies confirming these models due to the logistical constraints of tracking a forest as it develops over decades or centuries. Here we use a long-term dataset spanning >50 years of third-growth forest development in experimental forest gaps within a second-growth northern temperate hardwood forest to examine the influence of gap position on the structure and composition of regeneration and confirm past models of forest succession. We document key stand development patterns in even-aged mixed-species stands, including vertical stratification and shifts in diameter distributions. While the stands have shifted into a classic inverse-J diameter distribution, these shifts are due to a combination of unimodal distributions of the shade-intolerant and intermediate guilds that grow in mean diameter over time with consistently high densities of small individuals from the shade-tolerant guild. The greatest vertical stratification between the shade-intolerant and intermediate guilds occurred 30 years after clearing but then disappeared as the intermediate guild increased in height. In contrast, the separation in strata between the shade-tolerant species and the more intolerant guilds only increased over time. Finally, we found that gap position had a significant effect on productivity. The basal area, height, and species richness (in this case, an indicator of canopy stratification) were best modeled with negative quadratic regressions with highest values in gap centers and lower values towards the edges of gaps. This pattern aligns more with expected moisture and nutrient availability within gaps than expected solar radiation in the northern hemisphere, suggesting that moisture and nutrient availability may play a greater role than is often acknowledged in temperate forest types. Furthermore, these relationships were readily apparent only after 30 years yet became more accentuated with time, highlighting the importance of long-term research on successional dynamics for management. Ultimately our results document a clear successional trajectory in these gaps that follow expectations for even-aged mixed-species stand development where a full suite of species leads to rolling periods of dominance based on species functional guilds, with early successional species replaced by late-successional species in the progression of succession.