This observational retrospective cohort study provides benchmarking information on recent sow productivity development in Finnish herds. It focuses on parity cycle specific trends in sow removal patterns, and especially on the role of litter performance (size and piglet survival) in sow removal. In addition, the generated models offer a tool for calculating sow removal risks in any period, which could be used in economic and other simulation models. The data used in the study pool information of sows starting the same parity cycle (1 through 8) over the enrollment period of July 1st, 2013 through June 30th, 2014 and followed until the end of the study period (December 31st, 2014), and their performance histories across their entire previous productive life. Out of 71,512 individual sow parity cycle observations from the first to the eighth, 15,128 ended up in removal. Average litter sizes exceeded 13 piglets born in total in all of the most recent farrowings. Yet, even larger litter sizes were favored by the implemented culling policies, as sows having medium and large early life litters had lower risks of removal compared to those with the smallest litters, particularly in younger animals. With regard to piglets born just prior to removal, the smallest litter sizes were associated with the greatest culling risk for sows of that particular parity. In addition, having more than one stillborn piglet in the first and second litter put the sow at higher risk of being removed in all but the last (sixth through eighth) of the studied parity cycles. Investigation of removal patterns revealed a negative linear relationship between parity count and the mean days from farrowing to removal. More specifically, the median (mean) times to removal varied across the parity cycles from 62 (72) in the first to 34 days in both the seventh and eighth (47 and 42, respectively). Moreover, one in every six sows was removed within the first and second parity cycle. The findings especially in the earliest cycles may be a reflection of removal decisions not made according to any clear and pre-determined policy, or of biological issues that prevent farmers from firmly adhering to their policy. Quantitative performance should be linked to overall system functionality and profitability while taking animal welfare into consideration in identifying opportunities to improve herd parity structure and future farm success.