Reed canary grass (RCG, Phalaris arundinacea L.) is a suitable energy crop for cultivation in northern peatlands. However, the atmospheric impact of RCG cultivation as influenced by harvest frequency and fertilization is not clear. Here, we compared the biomass yield and greenhouse gas (GHG) balance for RCG cultivation in peatlands affected by cutting frequency and fertilizer managements. The managements included one-cut (OC) and two-cut (TC) systems that were either fertilized (TC-F) or unfertilized (TC-U) after the first cut in summer. Biomass yield of OC, TC-F and TC-U were 12, 16 and 11 Mg dry biomass per hectare per year, respectively. GHG fluxes of CO2, N2O and CH4 were measured with closed chamber techniques in the period between first and second (final) harvest of the TC managements, i.e. from 15 June to 23 September 2011. In the GHG monitoring period of 100 days, all systems were net sources of CO2 corresponding to 64 ± 3, 217 ± 15 and 50 ± 23 g CO2-C m−2 (mean ± standard error, n = 3) from the OC, TC-F and TC-U systems, respectively. In the same period, emissions of N2O from TC-F were ten times higher as compared to OC and TC-U. Emissions of CH4 were negligible from all systems. The TC systems could not improve the GHG balance during cultivation (271, 663 and 210 g CO2e-C m−2 emissions from the OC, TC-F and TC-U systems, respectively), but in a broader GHG life cycle perspective, the increased biomass yield by TC-F could replace more fossil fuel and offset at least some of the higher emissions from the system.