Suppose you have continuous $f_k$, $k=1,...,p$. Let $\psi(x) = \max_k f_k(x)$.
Choose $x$ and let $\epsilon>0$, and find $\delta>0$ such that
$-\epsilon < f_k(x)-f_k(y) < \epsilon$ for all $y \in B(x,\delta)$.
Rewrite as $f_k(x) < \epsilon + f_k(y) \le \epsilon +\psi(y)$, and now take the
$\max$ of the left hand side to get
$\psi(x) < \epsilon +\psi(y)$, or $\psi(x)-\psi(y) < \epsilon$.
Repeating for the other inequality gives
$-\epsilon < \psi(x)-\psi(y) < \epsilon$ for all $y \in B(x,\delta)$,
and so $\psi$ is continuous.