You can combine any number of algorithmic environments within your algorithm environment. The following MWE shows how you can do this to create a section called Inputs and one called Steps while still being able to break the algorithm over more than one page (with \algstore and \algrestore).
\documentclass{article}
\usepackage{algorithm}% http://ctan.org/pkg/algorithm
\usepackage{algpseudocode}% http://ctan.org/pkg/algorithmicx
\usepackage{caption}% http://ctan.org/pkg/caption
\begin{document}
\begin{algorithm}
\caption{My algorithm}
\textbf{Inputs}% Inputs section
\begin{algorithmic}[1]
\State Input $A$
\State Input $B$
\State Input $C$
\end{algorithmic}
\bigskip
\textbf{Steps}% Steps section
\begin{algorithmic}[1]
\Procedure{Euclid}{$a,b$}\Comment{The g.c.d. of a and b}
\State $r\gets a\bmod b$
\While{$r\not=0$}\Comment{We have the answer if r is 0}
\State $a\gets b$
\State $b\gets r$
\algstore{myalg}
\end{algorithmic}
\end{algorithm}
\clearpage
\begin{algorithm}
\ContinuedFloat
\caption{My algorithm (continued)}
\begin{algorithmic}
\algrestore{myalg}
\State $r\gets a\bmod b$
\EndWhile\label{euclidendwhile}
\State \textbf{return} $b$\Comment{The gcd is b}
\EndProcedure
\end{algorithmic}
\end{algorithm}
\end{document}
If you're not concerned about page breaking, I'd just set this in a tabularx. It's all about the column specification:
\documentclass{article}
\usepackage{tabularx}% Loads the array package
\begin{document}
\noindent
\begin{tabularx}{\textwidth}{
@{\hspace{1.5em}}% Space for left bullet
>{\leavevmode\llap{\textbullet~}\raggedright}% Left bullet + formatting of column
X% Left column specification
@{\quad\hspace{1.5em}}% Space between columns + right bullet space
>{\leavevmode\llap{\textbullet~}\raggedright\arraybackslash}% Right bullet + formatting of column
X% Right column specification
@{}% No column space on right
}
Mary had a little lamb, &
Her fleece was white as snow. And everywhere that Mary went, the lamb was sure to go. \\
Old MacDonald had a farm, &
E-I-E-I-O! And on this farm he had a cow, E-I-E-I-O! With a moo moo here, and a moo moo there.
\end{tabularx}
\end{document}
As mentioned, the column specification is the important thing here. Technically we only have 2 X-columns, each of which have a space of 1.5em on the left of it that is used to place a \textbullet inside. The look is similar to that of an itemize. The two columns have an additional \quad between them.
With the correct setup, the code for your construction should be fairly easy to read/interpret, and therefore be easily maintained.
Content can also be stretched vertically by adjusting \arraystretch accordingly. Headings and other tabular-related constructions are also easy to add, for example, the use of booktabs:
\documentclass{article}
\usepackage{tabularx,booktabs}
\newcommand{\fakeleft}{\multicolumn{1}{X}{}}
\let\fakeright\fakeleft
\begin{document}
\renewcommand{\arraystretch}{1.5}% Vertically stretch tabular constructions
\noindent
\begin{tabularx}{\textwidth}{
@{\hspace{1.5em}}% Space for left bullet
>{\leavevmode\llap{\textbullet~}\raggedright}% Left bullet + formatting of column
X% Left column specification
@{\quad\hspace{1.5em}}% Space between columns + right bullet space
>{\leavevmode\llap{\textbullet~}\raggedright\arraybackslash}% Right bullet + formatting of column
X% Right column specification
@{}% No column space on right
}
\toprule
\multicolumn{1}{X}{\centering\bfseries Left heading} &
\multicolumn{1}{X}{\centering\bfseries Right heading} \\
\midrule
Mary had a little lamb, &
Her fleece was white as snow. And everywhere that Mary went, the lamb was sure to go. \\
\fakeleft &
This is a right bullet without a left counterpart. \\
Old MacDonald had a farm, &
E-I-E-I-O! And on this farm he had a cow, E-I-E-I-O! With a moo moo here, and a moo moo there. \\
\bottomrule
\end{tabularx}
\end{document}
\fakeleft/\fakeright can be used interchangeably to produce a bullet without its opposite counterpart.
Best Answer
One solution with multicol