(1) Linear regression was used to relate snow accumulation during single and consecutive winters with white-tailed deer (Odocoileus virginianus) fawn:doe ratios, moose (Alces alces) twinning rates and calf:cow ratios, and annual changes in deer and moose populations. Significant relationships were found between snow accumulation during individual winters and these dependent variables during the following year. However, the strongest relationships were between the dependent variables and the sums of the snow accumulations over the previous three winters. The percentage of the variability explained was 36 to 51.

(2) Significant relationships were also found between winter vulnerability of moose calves and the sum of the snow accumulations in the current, and up to seven previous, winters, with about 49% of the variability explained.

(3) No relationship was found between wolf numbers and the above dependent variables.

(4) These relationships imply that winter influences on maternal nutrition can accumulate for several years and that this cumulative effect strongly determines fecundity and/or calf and fawn survivability. Although wolf (Canis lupus L.) predation is the main direct mortality agent on fawns and calves, wolf density itself appears to be secondary to winter weather in influencing the deer and moose populations.