diff --git a/bayesalpha/author_model.py b/bayesalpha/author_model.py
index 04a2371..4791b02 100644
--- a/bayesalpha/author_model.py
+++ b/bayesalpha/author_model.py
@@ -194,6 +194,8 @@ def fit_authors(data,
     sampler_type : str
         Whether to use Markov chain Monte Carlo or variational inference.
         Either 'mcmc' or 'vi'. Defaults to 'mcmc'.
+    sampler_args : dict
+        Additional parameters for `pm.sample`.
     save_data : bool
         Whether to store the dataset in the result object.
     seed : int
diff --git a/bayesalpha/author_plotting.py b/bayesalpha/author_plotting.py
new file mode 100644
index 0000000..2f59d60
--- /dev/null
+++ b/bayesalpha/author_plotting.py
@@ -0,0 +1,70 @@
+import numpy as np
+import warnings
+import functools
+try:
+    import matplotlib.pyplot as plt
+    import seaborn as sns
+    _has_mpl = True
+except ImportError:
+    warnings.warn('Could not import matplotlib: Plotting unavailable.')
+    _has_mpl = False
+    plt = None
+    sns = None
+
+
+def _require_mpl(func):
+    @functools.wraps(func)
+    def inner(*args, **kwargs):
+        if not _has_mpl:
+            raise RuntimeError('Matplotlib is unavailable.')
+        return func(*args, **kwargs)
+
+    return inner
+
+
+@_require_mpl
+def plot_trace(trace, varname, title=None, ax=None, **kwargs):
+    """
+    Plot samples from trace for a specific variable.
+
+    Parameters
+    ----------
+    trace : AuthorModelResult object
+        Result from ba.fit_authors
+    varname : str
+        Name of variable to plot. Must be one of ['mu_global', 'mu_author',
+        'mu_algo', 'alpha_author', 'alpha_algo']
+    title : str (optional)
+        Title of plot
+    ax : plt.axis object (optional)
+        Axis on which to plot
+    kwargs : dict (optional)
+        Additional keyword args to pass to sns.distplot
+    """
+
+    if varname not in ['mu_global', 'mu_author', 'mu_algo',
+                       'alpha_author', 'alpha_algo']:
+        raise ValueError("`varname` must be one of ['mu_global', 'mu_author', "
+                         "'mu_algo', 'alpha_author', 'alpha_algo']")
+
+    if ax is None:
+        _, ax = plt.subplots(figsize=[12, 4])
+
+    for i in trace.trace[varname]['chain']:
+        if varname == 'mu_global':
+            sns.distplot(trace.trace['mu_global'].sel({'chain': i}).values,
+                         **kwargs)
+        else:
+            suffix = varname.split('_')[-1]  # Either 'author' or 'algo'
+            for j in trace.trace[varname][suffix]:
+                sns.distplot(trace.trace[varname].sel({'chain': i,
+                                                       suffix: j}).values,
+                             **kwargs)
+
+    if title:
+        ax.set_title(title)
+
+    plt.xlabel(varname)
+    plt.ylabel('Probability')
+
+    return ax
diff --git a/bayesalpha/plotting.py b/bayesalpha/returns_plotting.py
similarity index 100%
rename from bayesalpha/plotting.py
rename to bayesalpha/returns_plotting.py