1: \begin{abstract}

2: A version of the fundamental mean-square convergence theorem is proved for

3: stochastic differential equations (SDE) which coefficients are allowed to

4: grow polynomially at infinity and which satisfy a one-sided Lipschitz

5: condition. The theorem is illustrated on a number of particular numerical

6: methods, including a special balanced scheme and fully implicit methods.

7: Some numerical tests are presented.

8: 

9: \noindent \textbf{AMS 2000 subject classification. }Primary 60H35; secondary

10: 65C30, 60H10.

11: 

12: \noindent \textbf{Keywords}. SDEs with nonglobally Lipschitz coefficients,

13: numerical integration of SDEs in the mean-square sense; balanced methods;

14: fully implicit methods; strong convergence; almost sure convergence.

15: \end{abstract}