A recent study interpreted significant effects of source of cytoplasm as evidence for cytoplasmic inheritance. An alternative interpretation that observed results reflected residual additive genetic effects unaccounted for by statistical models used was tested by computer simulation of records over 60 yr for milk yield (h2=.30) and fat percentage (h2=.60) under an additive genetic model with no cytoplasmic effects. First lactation records for the last 30 yr were analyzed by least squares according to models that included effects of cytoplasmic source, generations to cytoplasmic source, herd, and year as well as effects of sire (model 1), sire and record of dam (model 2), Predicted Difference of sire (model 3), or Predicted Difference of sire and ½ Predicted Difference of maternal grandsire (model 4). Ten replicates were performed. Significant cytoplasmic effects were observed for all models in at least 8 of 10 replicates for milk yield and in all replicates for fat percentage. Average F ratios for cytoplasmic effects were 1.43, 1.25, 1.46, and 1.43 for milk yield and 2.00, 1.29, 2.04, and 1.96 for fat percentage for models 1, 2, 3, and 4. Cytoplasmic variances were 1.4 and 3.2% of residual variances for milk yield and fat percentage for model 1. Unaccounted for additive genetic effects can produce spurious cytoplasmic effects.