Mouse embryos at the late 1-cell and late 2-cell stages were labelled with [3H]adenosine for periods of up to 320 min during which the specific activity of the ATP pool was constant. The time course of the molar accumulation of adenosine was calculated for tRNA, high-molecular-weight poly(A)- RNA and poly(A) tails versus internal regions of poly(A)+ RNA. Most of the adenosine incorporation into tRNA is due to turnover of the 3'-terminal AMP but some new synthesis of tRNA also appears to take place in both 1-cell and 2-cell embryos at a rate of about 0.2 pg/embryo/h. In the poly(A)- RNA fraction, an unstable component which is assumed to be heterogeneous nuclear RNA is synthesized at a high rate and accumulates at a steady-state level of about 1.5 pg/embryo in the 1-cell embryo and about 3.0 pg/embryo in the 2-cell embryo. Both 1-cell and 2-cell embryos synthesize relatively stable heterogeneous poly(A)- RNA, assumed to be mRNA, at a rate of about 0.3 pg/embryo/h; 2-cell embryos also synthesize mature ribosomal RNA at a rate of about 0.4 pg/embryo/h. Internally labelled poly(A)+ RNA is synthesized at a low rate in the 1-cell embryo, about 0.045 pg/embryo/h, but the rate increases to about 0.2 pg/embryo/h by the 2-cell stage. A striking feature of the 1-cell embryo is the high rate of synthesis of poly(A) tails, about 2.5 X 10(6) tails/embryo/h of an average length of (A)43, due almost entirely to cytoplasmic polyadenylation. This and other evidence suggests a turnover of the poly(A)+ RNA population in 1-cell embryos as a result of polyadenylation of new RNA sequences and degradation of some of the pre-existing poly(A)+ RNA. In the 2-cell embryo, the rate of synthesis of poly(A) tails (average length (A)93) is estimated at about 0.8 X 10(6) tails/embryo/h and a significant fraction of poly(A) synthesis appears to be nuclear.