[1] Measurements of NO, NO 2 , O 3 , and ultraviolet irradiance made at Cape Norman, Newfoundland (51.6°N, 55.9°W), during February-April 1996 were used to examine peroxy radical photochemistry and the local photochemical ozone tendency. Deviations from NO-NO 2 -O 3 photostationary state (the PSSD method) were used to estimate peroxy radical (PO 2 ) mixing ratios. Potential biases in the PSSD method and the methods used to apply it in an unbiased manner are discussed. PO 2 levels reached 77 ± 8 pptv (midday mean ±2s of the mean), significantly higher than observed in previous studies at remote marine locations. [PO 2 ] was well correlated with (J O 1 D ) 1/2 , indicating that the dominant radical source was photolysis by ultraviolet radiation in the wavelength region of O 3 photolysis and the dominant sink of PO 2 was recombination. However, the variation of [PO 2 ] with () was significantly higher than expected at this site during spring. Significant radical production in addition to that from net O 3 photolysis, possibly related to snow photochemistry in the upwind region, is suggested as the cause. However, the measurements needed to test this hypothesis were not made in this study, and we suggest future HO x measurements in remote midlatitude, snow-covered regions. The net photochemical ozone production at this site during springtime was calculated to be minimal (generally small and negative, and always less than or equal to +0.1 ppbv/h). These small values indicate that in situ photochemical ozone production is negligible at this site during spring; if [HO 2 ]/[PO 2 ] is large, net ozone destruction may contribute up to $10% of mean ozone levels per day.