The objective of this study was to obtain pilot data on which to judge the feasibility and sample size needed for a future comparative-effectiveness trial of platelet transfusions in the NICU. We conducted a limited-scope pilot trial in which neonates were randomized to receive platelet transfusions based on platelet mass vs. platelet count, using preset "transfusion-trigger" values. Analysis included parental consent rate, number of platelet transfusions given, bleeding episodes recorded, and mortality rate. Statistical analysis included ANOVA and Chi-square. A convenience sample of 30 were randomized; 15 per group. No differences were found between groups in gestational age, birth weight, race, gender or clinical diagnoses. The study consent rate was 52% (30/58). No differences were found in number of platelet transfusions received, bleeding episodes, or mortality. Lack of a trend in transfusion-reduction resulted in inability to estimate the number needed in a future comparative-effectiveness trial. Using platelet mass, rather than platelet count, for a NICU platelet transfusion trigger is feasible. However, any future comparative-effectiveness trial, testing the hypothesis that a platelet mass-based trigger reduces the transfusion rate will likely require a very large sample size.