Are they both 1 item (i.e., 1 likert item measuring fit, and 1 item measuring intentions)? Or are they both scales (i.e., both multiple items that together comprise 2 scales)?
In case it is just 1 item, they are ordinal variables and you need spearman rho correlation.
If they are both scales and you meet the assumptions for correlation, you can do correlation.
Since you only have 1 predictor variable and 1 outcome variable, the results of a regression and correlation will be the same, so it doesn't matter which one you choose.
If I somehow misunderstood your design, please provide more info on the variables (not the content but the variables, e.g. what type of variables they are (scale, ordinal... etc).