An enzyme packed microreactor was compared with a batch reactor system to determine effects of reaction water content on immobilized Candida antarctica Lipase B (Novozyme 435, N435) reusability, efficiency of end-group functionalization, apparent monomer conversion rate constants (k(app)), average molar mass (M-n), and leaching of Candida antarctica Lipase B (CALB) from Lewatit macroporous beads. Conversion of epsilon-caprolactone to poly(epsilon-caprolactone), PCL, using benzyl alcohol for end-functionalization, was the model system for studies conducted herein. The apparent rate constant in microreactor (k(app) = 0.027 s(-1)) was 27 times larger than in batch reactor (k(app) = 0.001 s(-1)). Furthermore, in microreactor, the M-n vs conversion plots for "dry" and "water saturated" conditions were similar. In contrast, under "water saturated" conditions in a batch reactor, M-n is much lower. Moreover, at both high and low water content, higher end-group functionalization was achieved for polymerizations in microreactor (0.75 to >= 0.98) as compared to batch reactor. (0.2). Also, microreactors run for 30 cycles under "dry" or "water saturated" conditions gave product where the fraction of benzyl ester groups on chains remains high, between 80 and 90%. In contrast, in the "water saturated" batch system, the fraction of benzyl ester terminal groups remains at about 0.30 throughout all 11 reaction cycles. These results led to the conclusion that the microreactor design results in effectively "dry" conditions even when reactants are "water saturated". CALB leaching during 7, 18, and 25 reaction cycles in microreactor steadily increased from 10 to 15 and 22%. However, at 30 cycles, CALB leaching disproportionaly increased to 76% without apparent physical deterioration of beads. Comparative results in batch reactors for all experiments above are reported and discussed.