The correspondence between R^2 or proportion of variance explained and model fit is not quite the same in multivariate models as it is in univariate ones. The high proportion of variance explained means that a very large amount of the variance in Y depends on your predictors in your model. The poor fit means that your model does a poor job of representing all of the relationships between the entire set of variables, including your predictors.
If you can/feel comfortable, post your code and data (either in whole, in part, or a simulation with similar characteristics) so we can get a better sense of what you're doing and be able to help you more.
The correspondence between R^2 or proportion of variance explained and model fit is not quite the same in multivariate models as it is in univariate ones. The high proportion of variance explained means that a very large amount of the variance in Y depends on your predictors in your model. The poor fit means that your model does a poor job of representing all of the relationships between the entire set of variables, including your predictors.
If you can/feel comfortable, post your code and data (either in whole, in part, or a simulation with similar characteristics) so we can get a better sense of what you're doing and be able to help you more.
ryne